Feed
AutoSaddler: Automatic Harness Optimization with Durable Updates from Agent Execution Traces
An AI coding agent debugs another agent's failures and patches its wrapper code, raising success rates.
9.0
pp
Pass@1 gain on GAIA2
Similar gains on coding and terminal benchmarks, without manual harness tuning
How it works
Harnesses are hand-tuned
Agents fail on long tasks. The prompts, tools and control code around them are tuned slowly by hand.
Debug traces, patch code
An agent reads failed runs and the harness source. It writes typed patches and keeps those that help unseen tasks.
Better agents, fewer runs
Test success rose 9 to 10 points on three benchmarks. It beat GEPA and Meta-Harness using fewer task executions.
From the paper